Discussion Forum

Reflections on overheating metros Aug 2019

Calvin R Barrows, BSc (Hons), CEng, MICE
As summer temperatures rise, concern rises also over the rising temperatures in the underground environment of overcrowded metro systems that include above and below ground sections and have no sub-surface cooling systems. Following on from his previous papers, Calvin Barrows, makes the case for painting trains with high-performance solar reflective paint and fitting low emissivity glass windows as a cost effective method of addressing the issues retrospectively in older metro systems.

There are many heat sources in metro systems and, while it is only those that are seasonal under consideration in this paper, it is worth noting the differences in operations during winter and summer. In cooler weather, the combined effect of the non-seasonal heat sources is insufficient to heat trains to a comfortable level for passengers without additional heating systems in trains. In summer there are two heat sources that increase temperatures in trains: the outside ambient air temperature and solar irradiation. The former is an important and notable seasonal variant and, while ambient air temperature varies throughout the daily cycle, it does not dramatically change over a 30 minute period. It cannot, therefore, explain the cause of the considerable increase in train temperatures on the surface often achieved over that same time period, primarily due to the role of solar irradiation.

Even though we all experience the effect of sunshine in summer, there does not appear to be more than a superficial appreciation of just how powerful solar irradiation can be. When the sun is at its zenith, direct sunlight at earth surface is about 1050 W/m2. This is supported by monitored and observational evidence. The power of the sun is also affected by the angle of incidence, which reduces its intensity when it is at shallower angles. For example, at a 45° angle of incidence, although solar radiation can cover a 40% greater area, it is then 30% less intense than when at its maximum angle of incidence of 90°. Consequently, how this changing intensity might affect trains also needs consideration.

Over heated trains adds to the misery of over crowded underground metro systems
Over heated trains adds to the misery of over crowded underground metro systems

Professors Piercarlo Romagnoni and Fabio Peron of the Università Iuav di Venezia produced a factsheet which examines the temperature impact of solar radiation on thermal insulation materials for roofing applications. Although some train roofs are not insulated, this factsheet gives an indication of the potential external skin temperatures. One test they undertook was on a roof sheet formed of a curved sandwich panel:

  • The external surface was aluminium metal sheet of 7/10mm thick and painted red;
  • The insulation was a 40mm thick layer of polyurethane foam;
  • The internal surface was a galvanized, corrugated metal sheet of 4/10mm thick.

While this is not exactly how a train roof is constructed, the external skin of a train roof would absorb similar levels of solar irradiation and reach similar temperatures. The maximum temperatures on curved roofs in more moderate climate zones can reach 67°C. Summer ambient air temperatures can reach 40°C, so the surface temperature of the aluminium external skin of this sandwiched panel is potentially around 25°C above ambient, which is akin to the difference between train rail temperatures and ambient air temperatures.

The next consideration is, how much of the train should be coated with solar reflective paint. Some have focused on the roof. However, focusing on the roof alone will limit the potential benefits. The average carriage size is 2.9m wide x about 2.5m high. Taking that as a basis, when the sun is at its zenith, it produces 1050 W/m2 x 2.9m/m length of carriage or 3045 W/m (3.045 kW/m) on its roof. With the sun at a 45° angle of incidence, and with the sun on both the roof and one or other side of the train, there is approximately 2.5m + 2.9m of train body exposed to the sun. At this angle the sun intensity is 1050 W/m2 x 70% (30% less than at its zenith) = 735 W/m2 but this is now shining on 5.4m/m length of train body. In this scenario then the sun produces 735 W/m2 x 5.4m/m length of carriage or 3969 W/m (3.969 kW/m).

The 1972 British Rail Research calculation had previously raised a concern over how solar irradiation is calculated and calculated the effect as being 7.8kW per trailer car or approximately 0.5kW/m length of carriage, which is clearly off the mark for the following reasons.

  • Train saloon heaters for winter operation produce around 15kW of heat per carriage, approximately twice the 1972 claimed power of the summer sun.
  • Even in low ambient winter temperatures people can be pleasantly comfortable when in the winter sunshine.
  • Recent reports by Transport for London about a trial investigation on the Central Line of the London Underground system compared the levels of solar heat gain through train windows fitted with tinted film to those without, and the extent to which untreated windows contributed to the elevated temperature in the carriages, in open sections of track, during the hottest portion of the day. A significant temperature difference was recorded. The carriage with tinted glass windows recorded temperatures of up to 4.5°C cooler. The limited, low performance measures implemented by the year 2015 would only have delivered these notable improvements if the effect of solar irradiation were significant, which it clearly is.

Sources of heat generation of underground metro trains and systems
Sources of heat generation of underground metro trains and systems

If we now consider the benefits of adding high-performance solar reflective paint on train bodies to reflect solar irradiation, it would have the effect to:

  • Reduce internal train temperatures in summer;
  • Reduce the required size of train air conditioning systems, delivering greater payload and reduced weight;
  • Reduce the in-train air conditioning running costs;
  • Improve the passenger experience in terms of comfort; and importantly,
  • Reduce the safety risk to passengers in a stalled train event.

Similarly, in the context of reducing tunnel heat, reflecting the solar irradiation when travelling on the surface in summer would:

  • Reduce the external skin temperature of the train which in turn would reduce the absorbed heat load into the train body and the heat emitted from the train body into the tunnel;
  • Reduce the internal temperature of the train, and the heat being discharged from the train air conditioning system (if fitted), on entry into the tunnel; and
  • Subsequently reduce the heat transferred across the network from hotter lines to cooler lines by the pressure/suction wave caused by the moving trains, while satisfying the laws of thermodynamics that heat will go to cold.

The benefits extend further. With solar reflective paint on trains, the operation of adequate and comprehensive cooling systems in underground environments may become less crucial, be under-run, or become fully or partially redundant. For underground metro networks without cooling systems, solar reflective paint may obviate the need to install them at all.

As obvious as all the above may be, testing still needs to be carried out. While it is simple to compare the reductions in internal train temperatures, assessing accurately the cumulative reduction in tunnel temperatures, as each train is treated with solar reflective paint, is less straightforward. Testing should involve continuous temperature measurements of all the relevant train temperatures, along its entire route, and be related to time, location and external ambient temperature. There are several additional points that will need careful consideration in order to quantify accurately the collective benefits in respect of reduced tunnel temperatures. The sooner these relevant train temperature readings can commence the sooner a reliable baseline can be established.

Undoubtedly the optimum test would be to treat all the trains on one line, and I would suggest, the most appropriate line would be the Central Line of the London Underground metro network, bearing in mind the 40°C+ temperatures experienced on the trains during the summer of 2018 and which reached 44°C in heatwave of this summer (2019) in Central Line trains travelling on the surface track section. Doing this optimum test would be the only way to measure, rather than theoretically predict, the overall reduction of tunnel temperatures. It would not be deemed sensible to treat all the trains without some evidence of the benefit, however, the limitations of data from a less comprehensive test will need to be fully understood and well thought through. Such a comprehensive test would also avoid skewing the results with such elements as the carry-in heat of untreated trains being transferred within the tunnel to the treated trains.

A London Underground train that could reduce the heat carried into the underground environment if painted with reflective paint and fitted with low emissivity glass windows
A London Underground train that could reduce the heat carried into the underground environment if painted with reflective paint and fitted with low emissivity glass windows

Monitoring the winter temperatures also would be useful to compare external train skin temperatures with those in the summer. There would seem little to be gained by monitoring internal saloon temperatures with their heaters operating, however, it is yet another legitimate cyclical heat source to be considered within the seasonal differential and would be useful, provided it is interpreted appropriately. Unlike the practice to date, all possible heat sources should be accounted for robustly.

Monitoring the external skin of a train might best be achieved with sensors on the internal face of the external skin and suitably insulated from the train’s internal space. This will mitigate the inaccuracies caused by unrepresentative cooling effects from, for example, local eddy currents caused by turbulent airflows. Monitoring of undercarriages and bogies would need some serious thought, in order to be able to identify the effects of the various heat sources operating on these.

With every additional treated train, first on the Central Line and then throughout the network, the cumulative reductions on tunnel temperatures, and thus sub-surface network temperatures, would become significant. The fullest benefits will only become apparent when all lines have had all their trains treated.

Whilst the foregoing looks at the general solar irradiation effect on the passenger compartment of the train, it should not be overlooked that the undercarriage (bogies) will also be affected. Traction and braking provide two non-seasonal heat sources acting on the bogies. However, there are three seasonal affects that need to be considered. Overlooking these may have skewed the previous perception of the heat load from braking. The three seasonal affects are:

  • The considerable mass of steel in the bogies and wheels are also affected when the sun is striking one or other side of the train;
  • The heat being emitted/transferred from the rails, which in the summer sun can often reach 20°C above ambient; and
  • The solar irradiated sleepers and ballast, which will also absorb the sun’s heat reflected to the underside of the train and bogies.

These additional heat loads are less likely to have a significant effect on the passenger compartments when overground, but will add to the carry-in heat within the tunnels. It therefore would make total sense to treat the exposed areas of undercarriage and wheels with the high-performance solar reflective paint. Recently Railtrack, which owns the nationwide rail infrastructure in the UK, has been painting non-wearing exposed surfaces of rails with white paint to mitigate rail buckling. Implementing this on the London Underground network would reduce the heat absorbed by the rails and consequently what they emit as the trains pass over them.

Glass window considerations

In considering the issues raised, painting the train body and undercarriage in this way is still only part of the solution, albeit a significant one. In addition, there is the matter of the windows. The effect of solar gain through glass windows is well known. The next logical step would be to fit low emissivity glass windows in all trains. Given the high percentage of window glass in the body of a carriage, using highly reflective glass would prevent the absorption of external solar heat while the train is on the surface (as with solar reflective paint) to:

  • Reduce the internal carriage temperatures; and
  • Reduce the external glass surface temperature to in turn, reduce the carry-in heat being emitted as trains pass into the underground.

Treatment with solar reflective measures would probably be best undertaken in stages:

  • Establish the baseline for tunnel temperatures with all trains being untreated;
  • Treat one train initially, including the undercarriage, with high-performance solar reflective paint and low emissivity windows;
  • Compare the reduction in heat absorbed by this treated train with the baseline data;
  • Extrapolate the above results to get a sense of the cumulative reductions in tunnel temperatures 
that can be achieved once all trains are treated.

Clearly treating only one train will have a negligible effect on tunnel temperatures but treating all the rolling stock of the line will achieve a significant reduction in tunnel temperatures across the treated line. Some insight as to the consequences of heat emittance on the treated train from the untreated trains in the tunnel would also be useful, since such a transfer of heat would result in an under-estimation of the ultimate benefits. By virtue of an inefficient transfer mechanism, the effect of overheated tunnel air on the treated train skin may not be of great significance.

Of course, when the temperature of one line (the hottest line) is reduced significantly, other lines would now be relatively hotter, as, by the laws of thermodynamics, there would be a transfer of heat from untreated lines to the now cooler line. As a result of the interdependencies of one line with the others, after the completion of the preceding hottest line, a decision on the treatment sequence of the other lines would be required.

In conclusion, a metro that has no surface lines would not benefit from having its trains painted with solar reflective paint or being fitted with low emissivity glass windows, providing the stabling of trains is properly managed. Conversely, metros with surface and sub-surface lines, and indeed overground trains, would benefit greatly from having their rolling stock painted with solar reflective paint and windows fitted with highly reflective glass.

Furthermore, in an era of heightened and pressing environmental concern, the application of solar reflective paint and the incorporation of highly reflective glass windows is cost-effective, particularly when compared to the capital cost of installing mechanical cooling solutions. In the longer term it would deliver real on-going reductions in maintenance and energy running costs.

As a final thought, of all the mitigation works implemented under the Cooling the Tube Project sponsored by Transport for London, the limited solar reflective foil and film treatment of Central Line trains, that reduced the internal temperature of the carriages by some 4°C, proved by far the most successful. If my information is correct, and I believe it is, the implemented Green Park Station Ground Source Cooling Project by London Underground was tendered at about £12 million and the final costs were recorded at £25 million, and it would have made little or no difference to train temperatures.

For that amount of money, an lot of high-performance solar reflective paint and low emissivity glass windows could be applied and fitted. The solar reflective paint and low emissivity glass windows would reduce the amount of heat absorbed by trains travelling on the surface and enabling them to jettison this heat continually along the overground route before they reach and enter the underground. Think of it as sun cream for trains!


Thanks are extended to fellow engineers Brian Farley, Philip Hargrave and John Newman, for their input and questioning approach, which added new perspectives on the many facets at play.


Add your comment

Thank you for taking the time to share your thoughts and comments. You share in the wider tunnelling community, so please keep your comments smart and civil. Don't attack other readers personally, and keep your language professional.
In case of an error submitting Feedback, copy and send the text to Feedback@TunnelTalk.com
Name :

Date :

Email :

Phone No :

   Security Image Refresh
Enter the security code :
No spaces, case-sensitive